Saved in:
Bibliographic Details
Main Authors: Prasai, Suraj, Du, Mengnan, Zhang, Ying, Yang, Fan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.03878
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914140194340864
author Prasai, Suraj
Du, Mengnan
Zhang, Ying
Yang, Fan
author_facet Prasai, Suraj
Du, Mengnan
Zhang, Ying
Yang, Fan
contents We develop KnowThyself, an agentic assistant that advances large language model (LLM) interpretability. Existing tools provide useful insights but remain fragmented and code-intensive. KnowThyself consolidates these capabilities into a chat-based interface, where users can upload models, pose natural language questions, and obtain interactive visualizations with guided explanations. At its core, an orchestrator LLM first reformulates user queries, an agent router further directs them to specialized modules, and the outputs are finally contextualized into coherent explanations. This design lowers technical barriers and provides an extensible platform for LLM inspection. By embedding the whole process into a conversational workflow, KnowThyself offers a robust foundation for accessible LLM interpretability.
format Preprint
id arxiv_https___arxiv_org_abs_2511_03878
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle KnowThyself: An Agentic Assistant for LLM Interpretability
Prasai, Suraj
Du, Mengnan
Zhang, Ying
Yang, Fan
Artificial Intelligence
Information Retrieval
Machine Learning
Multiagent Systems
I.2.7; I.2.0
We develop KnowThyself, an agentic assistant that advances large language model (LLM) interpretability. Existing tools provide useful insights but remain fragmented and code-intensive. KnowThyself consolidates these capabilities into a chat-based interface, where users can upload models, pose natural language questions, and obtain interactive visualizations with guided explanations. At its core, an orchestrator LLM first reformulates user queries, an agent router further directs them to specialized modules, and the outputs are finally contextualized into coherent explanations. This design lowers technical barriers and provides an extensible platform for LLM inspection. By embedding the whole process into a conversational workflow, KnowThyself offers a robust foundation for accessible LLM interpretability.
title KnowThyself: An Agentic Assistant for LLM Interpretability
topic Artificial Intelligence
Information Retrieval
Machine Learning
Multiagent Systems
I.2.7; I.2.0
url https://arxiv.org/abs/2511.03878